Experimental Characterization of Carbon Nanotube Cold...

American Institute of Aeronautics and Astronautics

1

Experimental Characterization of Carbon Nanotube Field

Emission Cathode Lifetime

Logan T. Williams1 and Mitchell L. R. Walker

2

High-Power Electric Propulsion Laboratory, Georgia Institute of Technology, Atlanta, Georgia, 30332

Victor S. Kumsomboone3 and W. Jud Ready

4

Georgia Tech Research Institute, Atlanta, Georgia, 30332

There is interest in the use of carbon nanotubes (CNTs) to create a field emission cathode

for small satellite applications and the neutralization of exhaust plumes of low-power electric

propulsion devices since field emission cathodes do not require a gas flow to operate. As a

part of the cathode’s development, the current emission output over the lifetime of the

cathode must be determined. The Georgia Tech Research Institute and the Georgia Tech

High-Power Electric Propulsion Laboratory have fabricated multiple field emission cathodes

that consist of multi-walled CNT arrays. Seven cathodes are characterized at pressures

below 10-5 Torr at constant voltage between the CNTs and the gate until failure occurs. The

maximum current density observed is 9.08 mA/cm2, the maximum power density is 9.08

W/cm2, and the maximum life-span is 368 hours. The behavior of the cathode current

emission consists of oscillations and sudden shifts thought to be caused by CNT interactions.

Resistive heating is thought to be the primary cause for failure.

I. Introduction

HE use of carbon nanotubes (CNTs) for field emission (FE) cathodes holds great potential as a general electron

emitter, especially in the area of electric propulsion (EP). There is a need for an efficient electron source to

neutralize the exhaust plume of low-power EP devices (< 500 W).1 Most thrusters operate in conjunction with

thermionic hollow cathodes, which require a gas flow in order to emit electrons. For a hollow cathode in use with a

200-W Hall effect thruster, which requires 0.8 A of cathode current, the propellant flow rate through the cathode is

typically 10% of the flow through the thruster itself.2 In addition, hollow cathodes require heating as well as a biased

keeper to ignite and initially sustain the cathode until the cathode is hot enough to self heat. A FE cathode only

requires an electric field for extraction of electrons. Since most spacecraft that use low-power EP systems have

limited power capacity and spacecraft mass allotment, this makes the simpler setup of the FE cathode a more

attractive alternative. Beyond the advantages from a systems design standpoint, FE cathodes can be very durable, as

CNTs are both chemically and mechanically robust.3 Recent work demonstrates the potential of FE cathodes not

only in the areas of low power EP, but also in the areas of propellant-less propulsion such as space tethers, and

spacecraft charge control.1,4

Field emission is the extraction of electrons from a conductive material through quantum tunneling. The

application of an external electric field lowers the potential barrier to the point that the transmission probability of an

electron becomes non-negligible. Field emission devices share the common feature that the emission sites are small

points which focus the electric field lines and increase the local electric field strength. This focusing effect allows

for low macroscopic electric field strengths (~1 V/µm) to enable electron emission. This focusing effect is termed

the field amplification factor. Past designs of FE cathodes consist of a base material, generally a doped n-type

semiconductor, with a layer of emitter of material on top. A conductive screen, termed the gate, placed above the

emitter is the extraction electrode. The electric field is applied between the emitter material and the gate. Spindt

developed the first FE cathodes in the late 1960’s which used metal cones as the emitter material. Subsequent FE

cathode designs use different emitter materials, including films made from diamond and CNTs.3,5-9 Past work by

1 Graduate Research Assistant, Aerospace Engineering, [email protected], AIAA Member.

2 Assistant Professor, Aerospace Engineering, [email protected], Senior AIAA Member.

3 Research Assistant, Georgia Tech Research Institute, [email protected].

4 Senior Research Engineer, Georgia Tech Research Institute, [email protected].

T

45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit2 - 5 August 2009, Denver, Colorado

AIAA 2009-5003

Copyright © 2009 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.


2

Williams et al studied the performance of vertically aligned CNT array FE cathodes.10 This current work studies the

current emission of seven CNT cathodes at constant voltage between the CNTs and the gate over time until emission

ceases.

II. Experimental Design

The CNT cathode follows the Spindt design pattern

where the cathode consists of a base, an insulator layer,

and a gate layer. Multi-walled CNT arrays are grown on

the base to serve as emitters. An electric field is generated

between the nanotubes and the gate to extract the electrons

from the CNT tips. The insulator electrically isolates the

base from the gate. Figure 1 shows a cut-away

representation of the cathode design.

The cathode fabrication process consists of successive layer deposition. The base is a wafer of n-type silicon

selected for electron donation. The insulator layer is made of silica (SiO2) and is thermally grown 10 microns thick

atop the silicon. The gate layer is formed by the deposition of a 200 nm thick layer of chromium with electron beam

evaporation. Silica and chromium are used as they do not inhibit CNT nucleation or growth.11

The chromium

surface is then covered with photoresist. A pattern of a repeated shape is made in the photoresist by

photolithography, which defines the shape of the nanotube arrays. The array patterns are defined by shape size (such

as length or diameter as appropriate) and pitch (the

separation distance between the centers of adjacent arrays).

After the pattern is developed, the chromium gate is etched

away via a standard chromium etch process. The SiO2

insulator is etched via reactive ion etching. Since the etch

rate differs between the two materials, more of the silica is

removed than the chromium. Thus, the gate layer extends

over the recesses of the insulator. The etched Cr and SiO2

allow a line-of-sight path for the deposition of an iron

catalyst layer directly on the silicon. The photoresist and

excess iron are removed via a stand liftoff process using

sonication in acetone. The CNTs are grown with chemical

vapor deposition (CVD) in a quartz furnace with methane,

acetylene, and hydrogen.12

Figure 2a shows images of circular CNT arrays from

one of the samples tested. Figure 2b shows a closer view of

the nanotube morphology obtained with a scanning

electron microscope (SEM). The arrays demonstrate a

degree of overall uniformity in shape, although there is a

slight variation between some arrays. Fringe nanotubes

exist at the edges of the arrays that are not vertically

aligned with the array. The individual nanotubes that make

up the arrays have an outer diameter of about 20 nm. The

density of the nanotubes within each array is about 10% by

area, which means that each nanotube is about 70 nm

apart. Despite the overall structure of the CNT arrays, the

individual nanotubes display a kinked vine-like

morphology and have a degree of anisotropy.12 A summary

of the geometry of the samples tested is outlined in Table

1.

The process used in this work created nanotubes that

extend beyond the gate, seen in Figure 2(a). The resultant

arrays have a height of about 50 microns, rather than the

expected 5-10 microns. Thus the resultant CNT samples

did not exactly match the original design. The design

illustrated in Figure 1 is an example of an intrinsic cathode,

Figure 1. Design configuration for CNCC.

Figure 2. Micrographs of CNT arrays (a) Non-

uniformity of arrays, (b) fringe CNTs near the

edge of an individual array.


3

where the nanotube arrays are entirely recessed within the

insulator. In order to generate an electric field that emits

electrons downstream of the cathode, an external gate is

suspended above the surface of the wafer. The electric

field is created between the external gate and the CNT

arrays rather than the chromium gate as outlined earlier.

This setup is referred to as the extrinsic configuration.

The wafer that contains the CNT arrays is mounted on

an aluminum body with Pelco colloidal silver liquid. The

gate is a 0.035 mm thick molybdenum grid with 0.14 mm

hexagonal holes with a 0.025 mm edge-to-edge separation.

The gate is mounted 1.3 mm above the surface of the

wafer. The cathode is then mounted in the vacuum

chamber.

III. Experimental Apparatus

A. Vacuum Facilities

All experiments are conducted in a 0.5 m diameter by

0.7 m tall stainless-steel bell jar vacuum system. Figure 3

shows a diagram of the bell jar vacuum system. The bell

jar is evacuated with a Varian VHS-6 diffusion pump

which is mechanically backed by an Alcatel 2033 SD

rotary vane pump. An uncooled optical baffle prevents oil

from back-streaming into the bell jar. The diffusion and

backing pumps have pumping speeds (for air) of 2400 l/s

and 8.33 l/s, respectively. Pressure in the bell jar is

measured by a Bayard Alpert 571 ion gauge in connection

with a SenTorr ion gauge controller with an overall error

of 20%,13

as well as two Varian 531 thermocouple pressure

transducers connected to Varian Model 801 thermocouple

gauge controllers.

B. Experimental Apparatus

The 1 cm by 1 cm silicon wafer which contains the

CNTs fits into a stainless steel mount which is mounted on

an aluminum bracket with four 2-56 size nylon machine

screws to maintain electrical isolation. Another aluminum

bracket is placed 2.5 cm opposite from the face of the

cathode body as an anode. Figure 4 shows the triode

configuration used to connect the cathode, gate, and anode.

The anode bias voltage, Va, is supplied by a Xantrex XPD

60-9 power supply. The cathode bias voltage, Vc, is

supplied by a Kepco BHK 2000-0.1MG power supply.

Both the anode and cathode power supplies share the same

ground. The currents to the gate, anode, and cathode are

measured across three 1-kΩ current shunts with a

combination of a Fluke multi-meter and an Agilent

34970A data acquisition unit (DAQ). The resistance value

has a tolerance of ±5%, the multi-meter has an uncertainty of 0.1%, and the DAQ has an uncertainty of 0.004% for

an overall uncertainty in the current density measurement of 5%.

D. Procedure

In order to study current emission, each sample is individually tested in the bell jar vacuum system. The

chamber is evacuated to a pressure below 1 x 10-5

Torr. The anode is biased to 50 V, and the cathode voltage is

progressively increased in 50 V steps at 5 minute intervals to at least 1000 V. The current is measured every five

Figure 3. Diagram of the bell jar vacuum system.

Table 1. Cathode CNT array patterns.

Sample Pattern

Shape

Pattern

Size

(µm)

Pitch

(µm)

CNT

Area

(cm2)

1 Triangle 4 16 0.0091

2 Diamond 8 64 0.0056

3 Diamond 2 8 0.0181

4 Star 8 32 0.0070

5 Star 50 150 0.0200

6 Diamond 1 8 0.0056

7 Square 4 16 0.0181

Figure 4. Cathode electrical schematic.


4

minutes until the cathode ceases to emit. The average current density is calculated as the current output divided by

the area of the CNT arrays.

IV. Experimental Results

Figure 5 shows the current density emitted by each cathode sample over time. One common characteristic

between all of the current emission of the samples over time is the existence of discontinuous jumps in the current

emission, particularly just before the sample fails. Some samples demonstrated multiple shifts in current emission,

such as samples 2, 3, 4, and 7. Table 2 shows a summary of the performance of the samples in terms of the cathode

lifetime, the maximum current density emitted, and the

maximum power density of the emitter.

Another major characteristic of the measured current

densities is that many has an oscillatory nature. Samples 1

and 6 both have a single dominant oscillation combined

with a slowly changing DC component, while samples 2

and 4 have a sum of several oscillations at different

frequencies combined with a DC component.

Figure 6 shows the current emission of each cathode

normalized by the maximum recorded value and plotted

against normalized time. This was done to demonstrate the

similarities between samples and to better see the shape of

trends in the emission current. Between samples 2, 4, and 5

each cathode experienced a sudden drop in current

emission at about 85% of the total lifetime.

All the cathodes demonstrate similar behavior of the

ratio of the anode current to the total current emitted from

the CNTs, called the transmission ratio. The higher the

transmission ratio, the greater the percentage of the

electrons emitted are available for an application. In each

case, the plots of current versus time for both the anode

and the gate currents have the same shape. If the current

collected by the anode drops, the current collected at the

gate experiences a similar drop. The offset between the

two, related to the transmission ratio, is not constant

a)

1.0

0.8

0.6

0.4

0.2

0.0No

rmal

ized

Cu

rren

t D

ensi

ty (

J/J m

ax)

1.00.80.60.40.20.0

Normalized Time (t/tmax)

1

6

b)

1.0

0.8

0.6

0.4

0.2

0.0No

rmal

ized

Cu

rren

t D

ensi

ty (

J/J m

ax)

1.00.80.60.40.20.0


2

4

5

c)

1.0

0.8

0.6

0.4

0.2

0.0No

rmal

ized

Cu

rren

t D

ensi

ty (

J/J m

ax)

1.00.80.60.40.20.0


3

7

Figure 6. Normalized current as a function of

normalized time density at constant voltage.

5

4

3

2

1

0

Cu

rren

t D

ensi

ty (

mA

/cm

2)

4003002001000

Time (hrs)

1

2

3

4

5

6

7

Figure 5. Summary of CNT cathode emission.

current density over time.


5

however. A common feature in the samples is that the

transmission ratio is initially around 50%, and then

decreases in an exponential fashion. Whenever the current

emission undergoes a shift, the transmission ratio likewise

shifts. Generally after each such shift the variance of the

transmission ratio increases. Figure 7 shows a plot that

illustrates both common features.

V. Discussion

There are two features of cathode operation over the

lifetime that are learned from the data: the failure

mechanism and the steady state behavior of the cathode.

The failure mechanism is most likely resistive heating.

There is ample evidence of thermal damage, and cathode

lifetime decreases as the cathode power density increases.

The behavior of the current emission over time is not a

function of the initial geometry of the CNT arrays. With

the composition of the CNTs being consistent, it is thought

that current switching caused by interactions between

nanotubes are the cause for the varying behavior between

samples.

A. Failure Mechanism

The first piece of evidence about the failure

mechanism is that many of the samples underwent shifts in

current emission at the same normalized time. While the

current density shifts did not always occur at the same

relative time, the fact that shifts overlapped the majority of

the time suggests that the phenomenon is rate dependent,

rather than a function of absolute time. This means that

nanotube failure is a result of some physical process within

the nanotubes. The data also shows that this process is

related to the power density applied through the emitter,

since cathode lifetime decreases as average power density

increases. This suggests resistive heating, which is the process where Ohmic heating from the current passing

through the nanotubes heat the tubes and enhance oxidative ablation, or outright vaporize the outer wall.14 A higher

power density causes a higher current flowing through a given number of nanotubes. Since individual nanotubes

have the same diameter, neglecting the slight variations in height between nanotubes, a higher power density would

result in a faster rise in temperature in the nanotubes. This in turn increases the rate of oxidative ablation and

vaporization. After each test the anode contains a film of carbon, which confirms that CNT material is vaporized and

ejected from the cathode during operation.

Another piece of evidence that suggests resistive heating is the presence of thermal damage on the face of the

emitters. Figure 8 shows a compilation of optical microscope images of sample 4 that show regions of scorching on

the right half of the wafer. Damage of this kind is present on every sample, which suggests that there are regions of

high thermal deposition. This supports the idea that resistive heating is the mechanism in question. The existence of

multiple scorch marks suggests that there are clusters of CNT arrays that act as primary emitters. As current is

emitted, heat is transferred from the CNT arrays to the surrounding material, causing the thermal damage. The

reason for multiple scorches is that as one region fails another activates, causing thermal damage in another location.

Resistive heating is not the only thermal event that occurs. One possibility is that resistive heating might lead to

arcing, as the vaporization of the nanotube walls from resistive heating might create a conductive path to the gate.

Figure 9 shows a region of sample 1 that has a solidified flow from melting the silicon substrate in the center of a

scorch mark. The high temperature needed to melt silicon (1420 ºC) suggests that arcing occurred at that point,

rather than some other phenomenon. Gröning et al determined that arcing in a CNT film cathode could create

currents as high as 8 x 105 mA/cm

2 in 200 ns pulses that could deposit 10

4 – 10

5 J/cm

3.15 Such an arc could cause the

damage seen in Figure 9, and a 200 ns pulse would not be registered by the DAQ unit. Such arcing potentially

16

14

12

10

8

6

4

2

0

Curr

ent

(µA

)

3002001000

Time (hours)

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Tran

smissio

n R

atio

Ianode

Igate

Ianode/Itotal

Figure 7. Transmission ratio of sample 7. Anode

voltage 50 V, cathode voltage 1200 V.

Table 2. Cathode Performance.

Sample Life

(hrs)

Max Current

Density

(mA/cm2)

Average

Power

(W/cm2)

1 250 2.10 1.45

2 83 4.83 2.73

3 249 0.50 0.10

4 117 9.08 2.52

5 145 0.95 0.27

6 113 0.44 0.53

7 368 1.54 0.21


6

occurs when vaporized CNT material creates a conductive

pathway to the gate.14 Work by Sowers shows that for the

case of a diamond film cathode, arcing decreased the

electric field required to extract electrons, which Sowers

attributed to a reshaping of the film to include more jagged

points. Whether this type of structural reshaping occurs in

CNT films is unknown. The shifts in current emission

could also be due to arcing damaging the emission sites,

which would result in the activation of other emission sites

that might not have as high a local field amplification

factor. Therefore an arc between the gate and the cathode

could cause large changes in current emission by either

altering the shape of the CNTs and increasing the field

amplification factor or by changing emission sites after an

arcing event. However, as there is no evidence of

widespread arcing, resistive heating is thought to be the

dominant mechanism for failure.

There are several alternative failure mechanisms that

were considered and rejected. Work by Bonard14 suggests

several different mechanisms by which CNT emitters may

fail. Of the various means explored by the literature, those

most applicable to the observed emission currents are wall

shedding of multi-walled nanotubes (MWNT) into single-

walled nanotubes (SWNT) and arcing between the

nanotubes and the gate. While it is tempting to consider

another of Bonard’s mechanisms, failure due to

mechanical stresses, it did not seem particularly feasible in

this situation. Failure caused by mechanical stresses has

generally been associated with CNT films, where the

nanotubes are not arranged in a given orientation and are

forced into a vertical position by the applied electric field.

In this case, the CNTs are grown in vertical arrays. While

the kinked morphology might introduce some strain as the

nanotubes straighten in the electric field, the stress would

be much lower than in CNT films. Additionally, the

cathodes operated at electric field intensities much lower

than what Bonard observed for failure (~1 V/µm versus 3

V/nm for Bonard).

Other hypothesized failure modes in the literature only

explain slight variations in the current emission. Some of

the basic failure modes for single-walled nanotubes, such as selective oxidation, ion bombardment, and field

evaporation, might be the cause for minor fluctuations in the current emission of the cathodes.18 While no

experiments have been done to tie such mechanisms to multi-walled nanotubes (MWNTs), it is a simple matter to

consider their impacts on MWNTs. Should an outer wall of a MWNT be weakened by an ion impact or from

oxidation, a section of the nanotube could be ejected and follow the electric field lines to the anode. The departure of

a section of a MWNT would disturb the local electric field lines and thus the current emission from the nearby

nanotubes. Likewise, current saturation and breakdown of the CNTs might be a factor.16 However, these

mechanisms would only describe moderate fluctuations in the current emission, and would particularly not explain

the large discontinuities in current emission seen in samples 2, 4, and 5 as that would imply a large number of

nanotubes are shedding the outer walls at the same time. Such an occurrence, while possible, would be unlikely

considering that it would require a very large number of active emission sites.

Most likely the length of the nanotubes is the ultimate cause for the failure of the cathodes. In terms of resistive

heating, an increase in nanotube length increases the resistance and thus the heat generated. The increased amount of

material also increases the likelihood of vaporized CNT material forming a conductive path to the gate and initiating

arcing. A refinement of the fabrication process to reduce the CNT height to the designed value should greatly reduce

the above processes and increase cathode life span.

Figure 8. Thermal damage on sample 4. Arrows

drawn point to the scorch marks.

Figure 9. Melted substrate on sample 1. Arrow

points to melted substrate.


7

B. Cathode Geometry and Current Switching

The geometric configuration of the arrays has no direct effect on the behavior of the current emission over time.

Work by Suh et al demonstrates that CNTs are sensitive to field screening effects for nanotubes as sparse as one per

50 µm2.17 For comparison, the samples tested have a nanotube density of one per 0.001 µm

2, and array densities

between one per 16 µm2

and one per 5625 µm2. The expectation is that the array geometry or structure may have

some impact on the current emission, as field screening effects vary with array size and pitch. However, some

samples with similar behavior of the current emission have completely different geometric characteristics. As an

example, sample 2 consists of arrays that individually have incomplete growth, yet shows similar behavior to sample

4, which has arrays that are overgrown. The only observable geometric correlation is that as the area occupied by the

nanotubes increases, the lifetime of the cathode increases. This makes sense, as the smaller the area, the greater the

power per unit area extracted from the cathode.

While the behavior of the current emission varies between cathodes, the behavior of the transmission ratio is

very similar between samples. It would be expected that the transmission ratio would be constant, as the geometry of

the gate does not change. However, not only is the transmission ratio not constant, it behaves the same way for each

sample. The decrease in the transmission ratio might is not due to the deposition of material on the grids, which

would physically decreasing the amount of free space for the electrons to pass through. The same gate is used for

each sample, and the transmission ratio each time would initially have a value between 0.5 and 0.6, which is higher

than some of the minimum values observed. This instead suggests some mechanism involving the CNTs. Similarly,

it is not understood why the transparency would change after any sudden shift in current emission. When this

occurs, the transmission ratio changes from an exponential decay in time to a linear change in time, if not constant.

As these traits were present in all samples, the phenomenon seems to be independent of CNT geometry.

It is possible that the varying behavior of the current emission is a result of interactions between CNTs. In

addition to the fact that the geometry has no effect on the behavior of the current emission, all the samples were

created using the same process, so individual CNT composition is the same. One possible explanation is that the

CNTs undergo current switching. Current switching is a sudden change in current emission caused by changes in the

emitting CNT tip. Current switching can be caused by motion of the nanotube, as different facets of the CNT

structure have slightly different work functions. Collins and Zettl18 show that current switching caused by

interactions between nanotubes could shift current emission by an order of magnitude in only a few seconds. Since

small perturbations could have a large effect on current emission, it is possible that the behavior of the current

emission between samples diverges as CNTs start to interact in a non-linear fashion. Such CNT interactions explain

the observed oscillations as an oscillation in the tip interactions between CNTs.

VI. Conclusion

Seven CNT cathodes fabricated from vertically aligned arrays are operated at constant voltage between the gate

and CNTs over time to provide current densities up to 4.24 mA/cm2 and power densities of up to 4.24 W/cm

2 for

lifetimes up to 368 hours. The current emission over time is characterized by oscillations and sudden shifts, with

resistive heating as the primary cause for failure. The behavior of the cathodes over time is independent of initial

CNT geometry and are thought to be determined by current switching caused by CNT interactions.

Acknowledgments

The authors of this paper would like to thank Prof. Lisa Pfefferlee of Yale University for technical assistance, the

Georgia Tech Aerospace Engineering Machine Shop for fabrication and hardware support, and the graduate students

of the High-Power Electric Propulsion Laboratory for additional assistance. This work is sponsored by DARPA per

contract number HR0011-07-C-0056.

References

1Mueller, J., Marrese, C., Wirz, R., Tajmar, M., Schein, J., Reinicke, R., Hruby, V., “JPL Micro-thrust Propulsion Activities,”

Proc. NanoTech, 2002. 2Hargus, W. A., and Reed, G., “The Air Force Clustered Hall Thruster Program” Proc. 38th AIAA/ASME/SAE/ASEE Joint

Propulsion Conf. and Exhibit, 2002. 3Gasdaka, C. J., Falkos, P., Hruby, V., Robin, M., Demmons, N., McCormick, R., Spence, D., Young, J., “Testing of Carbon

Nanotube Field Emission Cathodes,” 2004. 4Gilchrist, B. E., Jensen, K. L., Gallimore, A. D., Severns, J. G., “Space Based Applications for FEA Cathodes (FEAC),”

Materials Research Society Symposium Proceedings, Vol. 621, 2000.


8

5Spindt, C. A., Holland, C. E., Rosengreen, A., Brodie, I., “Field-Emitter Arrays for Vacuum Microelectronics,” IEEE

Transactions on Electron Devices, Vol. 38, No. 10, 1991, pp. 2355-2363. 6Agüero, V. M., Adamo, R. C., “Space Applications of Spindt Caathode Field,” 6th Spacecraft Charging Technology

Conference Proceedings, 2000, pp. 347-352. 7de Heer, W.A., A. Chatelain, and D. Ugarte, “Carbon nanotube field-emission electron source,” Science, Vol. 270, No.

5239, 1995, pp. 1179.

8Gamero-Castaño, M., Hruby, V., Falkos, P., Carnahan, D., Ondrusek, B., Lorents, D., “Electron Field Emission from Carbon

Nanotubes, and its Relevance in Space Applications,” Proc. 36th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 2000.

9Sowers, A. T., Ward, B. L., English, S. L., and Nemanich, R. J., “Field emission properties of nitrogen-doped diamond

films,” Journal of Applied Physics, Vol. 86, No. 7, 1999, pp. 3973-82. 10Williams, L., Kumsomboone, V. S., Ready, W. J., Walker, M. L. R., “Experimental Characterization of a Carbon Nanotube

Field Emission Cathode”, Proc. 44th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 2008.

11Hofmann, S., et al., “Direct growth of aligned carbon nanotube field emitter arrays onto plastic substrates,” Applied Physics

Letters, Vol. 83, No. 22, 2003 pp. 4661-4663.

12Turano, S. P., Ready, J., “Chemical Vapor Deposition Synthesis of Self-Aligned Carbon Nanotube Arrays,” Journal of

Electronic Materians, Vol. 35, No. 2, 2006 13Varian, Vacuum Measurement, In Varian, Inc., 2003 14Bonard J. and Klinke C., “Degradation and failure of carbon nanotube field emitters,” Physical Review B, Vol. 67, 2003. 15Gröning, O., Küttel, O. M., Schaller, I., Gröning, P., and Schlapbach, L., “Vacuum arc discharges preceding high electron

field emission from carbon films,” Applied Physics Letters, Vol. 69, No. 4, 1996, pp. 467-468. 16Collins P. G. et al, “Current saturation and electrical breakdown in multiwalled carbon nanotubes,” Physical Review

Letters, Vol. 86, No. 14, 2001, pp. 3128-3131. 17 Suh, J. S., et al., “Study of the field-screening effect of highly ordered carbon nanotube arrays,” Applied Physics Letters,

Vol. 80, No.13, 2002, pp. 2392-2394. 18Collins P. G. and Zettl A., “Unique characteristics of cold cathode carbon-nanotube-matrix field emitters,” Physical Review

B, Vol. 55, No. 15, 1997, pp. 9391-9399.

Experimental Characterization of Carbon Nanotube Cold...

Documents

Transcript of Experimental Characterization of Carbon Nanotube Cold...